Forecasting egg production curve with neural networks

ABSTRACT: The comparison between the real egg production curve and the graph proposed by management guidelines, aims towards continuous performance evaluation. The objectives of this study was to compare the capacity of curve fitting daily egg production of Lokhorst (LM), neural network multilayer perceptron (MP) and Jordan and Elman recurrent neural network (RNNJ and RNNE, respectively) for the prediction of the daily egg production in commercial laying hens. The models were fitted using 4650 data from 12 selected batches. The MP and LM models gave good fitting to the data, with correlation values greater than 0.95 and accounting for more than 95% of the variability in daily egg production. For the production forecast, MP was a technique with acceptable accuracy and less variation. The MP model can be recommended as a tool for fit and forecast of daily egg production curve in commercial hens.

• Autores: Galeano Vasco, Luis Fernando; Cerón Muñoz, Mario Fernando; Galvan, I.M.; Aler, R.
• Tipo de recurso: Article of investigation
• Fecha de publicación: 2018
• Institución: Universidad de Antioquia
• Repositorio: Repositorio UdeA
• Idioma: eng
• OAI Identifier: oai:bibliotecadigital.udea.edu.co:10495/32693
• Acceso en línea: https://hdl.handle.net/10495/32693
• Palabra clave: Modelos Teóricos; Models, Theoretical; Curvas de frecuencia; Frequency curves; Polinomios; Polynomials; Funciones; Functions; Avicultura; Aviculture; Producción de huevos; Egg production; http://aims.fao.org/aos/agrovoc/c_2498
• Rights: openAccess
• License: http://creativecommons.org/licenses/by-sa/2.5/co/